Larry Holder School of EECS Washington State University Artificial - - PowerPoint PPT Presentation

Larry Holder School of EECS Washington State University

1 Artificial Intelligence

} Weak AI

• Machines can act as if they were intelligent

} Strong AI

• Machines can actually be intelligent (i.e., think)

} Can we tell the difference? } Is even weak AI achievable? } Should we care about achieving strong AI? } Are there ethical implications?

Artificial Intelligence 2

} Turing Test

• Can the machine convince a human that

it is human via written English

} Loebner Prize

• en.wikipedia.org/wiki/Loebner_Prize

} AI XPRIZE (ai.xprize.org): $5M } Mitsuku (mitsuku.com)

Artificial Intelligence 3

Alan Turing (1912-1954)

The Singularity Is Near (2012)

} Disability

• But a machine can never…

– Beat a master at chess (ü) – Compose a symphony (~) – Laugh at a joke – Appreciate beauty – Fall in love

} Response

• Magenta Project (magenta.tensorflow.org)
• Engineer different approaches (planes vs. birds)
• If we can understand how humans do it…

Artificial Intelligence 4

} Mathematical objection

• Godel’s incompleteness theorem

– In any formal system there are true sentences that cannot be proven – “This sentence is not provable” is true, but not provable

} Response

• Formal systems are infinite, machines are finite
• Inability to prove obscure sentences not so bad
• Humans have limitations too

Artificial Intelligence 5

Kurt Godel 1906-1978

} Informality

• Human behavior too complex to model formally

} Response

• Usually assumes overly-simplistic models (e.g.,

propositional logic)

• Learning can augment the model

Artificial Intelligence 6

} Machine thinks like a human } How do we define human thinking?

• Machine has to know it passed the Turing test
• Consciousness argument

} Mental state = physical (brain) state } Mental state = physical state + ? } Arguments ill-defined } What is consciousness?

Artificial Intelligence 7

} Functionalists say "Yes"

• Brain maps inputs to outputs
• Can be modeled as a giant

lookup table

• Brain in a vat

} Naturalists say "No"

• Lookup tables are not intelligent
• Searle’s Chinese room argument

} Does achieving strong AI

matter?

Artificial Intelligence 8

Brain in a Vat Searle's Chinese Room

} Impact on economy: Losing jobs to automation } Lethal and autonomous robots } Surveillance and privacy } Data mining

Artificial Intelligence 9

“Eagle Eye” (2008) “Person of Interest” (2011-2016)

} AI responsibility

• Generally, human experts are responsible for

relying on AI decisions

• Autonomous AI liability falls to the human

designers

• Can an AI system be charged with a crime?

Artificial Intelligence 10

“I, Robot” (2004)

} Stephen Hawking (2014)

• “Success in creating AI would be the biggest event in

human history. Unfortunately, it might also be the last.”

} Bill Gates (2015)

• “I am in the camp that is concerned about super

intelligence.”

} Elon Musk (2017)

• “AI is a fundamental risk to the existence of human

civilisation.”

} Henry Kissinger (2018)

• “… whose culmination is a world relying on machines

ungoverned by ethical or philosophical norms.”

Artificial Intelligence Laboratory 11

1.

A robot may not injure a human being or, through inaction, allow a human being to come to harm.

2.

A robot must obey orders given it by human beings except where such orders would conflict with the First Law.

3.

A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

Artificial Intelligence Laboratory 12

Isaac Asimov 1920-1992

} Avoid Negative Side Effects

• How can we ensure that an AI system will not disturb its

environment in negative ways while pursuing its goals?

} Avoid Reward Hacking

• How can we avoid gaming of the reward function?

} Scalable Oversight

• How can we efficiently ensure that a given AI system respects

aspects of the objective that are too expensive to be frequently evaluated during training?

} Safe Exploration

• How do we ensure that an AI system doesn’t make exploratory

moves with very negative repercussions?

} Robustness to Distributional Shift

• How do we ensure that an AI system recognizes, and behaves

robustly, when it’s in an environment very different from its training environment?

Artificial Intelligence Laboratory 13

research.googleblog.com/2016/06/bringing-precision-to-ai-safety.html

} End of human race

• An unchecked AI system

makes a mistake

• Utility function has

undesired consequences

• Learning leads to

undesired behavior

• Singularity

} Friendly AI

Artificial Intelligence 14

“The Matrix” (1999) “Terminator 3: Rise of the Machines” (2003)

“Colossus: The Forbin

Project” (1970) “I, Robot” (2004) “Transcendence” (2014)

} Robot/AI rights

Artificial Intelligence 15

“Bicentennial Man” (1999) “A.I. Artificial Intelligence” (2001) “Ex Machina” (2015) “The Machine” (2013)

} Artificial Intelligence for the American People

• www.whitehouse.gov/ai/

Artificial Intelligence Laboratory 16

} Weak AI vs. Strong AI } Controlling AI } AI Laws } AI Rights } Human Future } Policy

Artificial Intelligence 17